H2O Danube3 500M
H2O.ai · dense · 0.5B parameters · 8,192 context
Parameters
0.5B
Context Window
8K tokens
Architecture
Dense
Best GPU
B200 SXM
Intelligence Brief
H2O Danube3 500M is a 0.5B parameter DENSE model from H2O.ai, featuring Grouped Query Attention (GQA) with 24 layers and 896 hidden dimensions. With a 8,192 token context window, it supports general text generation. For self-hosted inference, B200 SXM delivers optimal throughput at $4261/month.
Architecture Details
Memory Requirements
BF16 Weights
1.0 GB
FP8 Weights
0.5 GB
INT4 Weights
0.3 GB
GPU Compatibility Matrix
H2O Danube3 500M is compatible with 100% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
FP8 · 1 GPU · tensorrt-llm
83/100
score
Throughput
3.5K tok/s
Latency (ITL)
0.3ms
Est. TTFT
0ms
Cost/Month
$4261
Cost/M Tokens
$0.46
FP8 · 1 GPU · tensorrt-llm
83/100
score
Throughput
3.5K tok/s
Latency (ITL)
0.3ms
Est. TTFT
0ms
Cost/Month
$4271
Cost/M Tokens
$0.46
FP8 · 1 GPU · tensorrt-llm
83/100
score
Throughput
3.5K tok/s
Latency (ITL)
0.3ms
Est. TTFT
0ms
Cost/Month
$6169
Cost/M Tokens
$0.67
Deployment Options
API Deployment
No API pricing available
Single GPU
B200 SXM
$4261/mo
Min VRAM: 1 GB
Multi-GPU
B200 SXM
3.5K tok/s
Best available config
API Pricing Comparison
No API pricing data available for this model.
Performance Estimates
Throughput by GPU
VRAM Breakdown (B200 SXM, FP8)
Precision Impact
bf16
1.0 GB
weights/GPU
fp8
0.5 GB
weights/GPU
~3.5K tok/s
int4
0.3 GB
weights/GPU
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy H2O Danube3 500M
Self-Hosted Infrastructure
Similar Models
Qwen 2.5 0.5B
0.5B params · dense
Quality: 50
Cohere Embed English v3
0.5B params · dense
Quality: 50
from $0.10/M
InfoXLM Large
0.55B params · dense
Quality: 50
Multilingual E5 Large
0.56B params · dense
Quality: 50
BGE M3
0.568B params · dense
Quality: 50
from $0.01/M
Frequently Asked Questions
How much VRAM does H2O Danube3 500M need for inference?
H2O Danube3 500M requires approximately 1.0 GB of VRAM at BF16 precision, 0.5 GB at FP8, or 0.3 GB at INT4 quantization. Additional VRAM is needed for KV-cache (12288 bytes per token) and activations (~0.10 GB).
What is the best GPU for H2O Danube3 500M?
The top recommended GPU for H2O Danube3 500M is the B200 SXM using FP8 precision. It achieves approximately 3.5K tokens/sec at an estimated cost of $4261/month ($0.46/M tokens). Score: 83/100.
How much does H2O Danube3 500M inference cost?
H2O Danube3 500M inference costs vary by provider and GPU setup. Use our calculator for detailed cost estimates across all providers.